What are computer-use agents?
Understand the observe–plan–act loop, how GUI agents differ from browser automation, and which research resources help you get started.
Guides to computer-use agents
Practical reading guides with direct answers and primary research references. These guides help distinguish interface grounding, task execution, training data, and evaluation.
How to choose a GUI-agent benchmark
Compare desktop task execution, browser workflows, mobile environments, and GUI grounding, and preserve the evaluation conditions needed for meaningful comparisons.
Where to find training data for GUI agents
Distinguish grounding examples, interaction trajectories, screen understanding data, and benchmark tasks, with primary dataset sources and separate licensing checks.